Skip to main content

Full text of "Nigerian lizards of the genus Agama (Sauria: Agamidae)."

See other formats

\jP Biodiversity 

Bulletin of the British Museum (Natural History). 

London :BM(NH) 

17: http://www.biodiversitylibrary.0rg/item/l 9623 

Page(s): Title Page, Table of Contents, Page 65, Page 66, Page 67, Page 68, Page 69, Page 

70, Page 71 , Page 72, Page 73, Page 74, Page 75, Page 76, Page 77, Page 78, Page 79, Page 

80, Page 81 , Page 82, Page 83, Page 84, Page 85, Page 86, Page 87, Page 88, Page 89, Page 

90, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text 

Contributed by: Natural History Museum, London 
Sponsored by: Natural History Museum Library, London 

Generated 1 1 August 2009 2:30 PM 

This page intentionally left blank. 





VOL. 17 
1968 — 1969 




LONDON: 1973 






2 9 MAY ,m 





No. I. Notes on woodpeckers (Picidae). By D. Goodwin 

No. 2. A review of the iguanid lizard genera Uracentron and Strobilurus. 

By R. Etheridge 

No. 3. Nigerian lizards of the genus Agama (Sauna: Agamidae). By Alice 

G. C. Grandison (Pis. 1-6) 

No. 4. A revision of the amphipod genus 

Aoridae). By A. A. Myers (PI. i) 

us Costa (Gammaridea : 

No. 5. Convergence in the structure of the head and cuticle of Euckromadora 

species and apparently similar nematodes. By W. Grant Inglis 

No. 6. A nomenclatural index to "A history of the British marine Polyzoa'* 

by T. Hincks (1880) . By J. S. Ryland 

No. 7. The clupeoid fishes described by Bloch and Schneider. By P. J. P. 

Whitehead (Pis. 1-3) 

No. 8. Cyclophyllidean cestodes from birds in Borneo. By Michael D. B. 


No. 9. Type material of the families Lysianassidae, Stegocephalidae, 

Ampeliscidae and Haustoriidae (Crustacea: Amphipoda) in the 
collections of the British Museum (Natural History). By M. H. 

Thurston & Elizabeth Allen 











Index to Volume 17 


V '^ 

^^ Nov 






British Museum (Natural History) 


pp. 65-90 ; 6 Plates : 2 Text- figures 



Vol. 17 No. 3 


LONDON : 1968 


(NATURAL HISTORY), tnsmuted tn 1949, ts 
issued in five series corresponding to the Departments 
of the Museum, and an Historical series. 

Parts will appear at irregular intervals as they become 
ready. Volumes will contain about three or four 
hundred pages, and will not necessarily be completed 

one catenaar year. 

In 1965 a separate supplementary series of longer 
papers was instituted, numbered serially for each 

This paper is Vol. 17, No. 3 of the Zoological 
series. The abbreviated titles of periodicals cited 
follow those of the World List of Scientific Periodicals. 

World List abbreviation : 
Bull. Br. Mus. nat. Hist. (Zool.) 

Trustees of the British Museum (Natural History) 1968 


19 November, 1968 

Price Eighteen Shillings 




The external morphology, distribution and systematics of the species of Agama occurring in 
Nigeria are reviewed. The number and position of the dermal sense organs on the trunk scales 
are used to help clarify the status and relationships of the five species recognized. Agama 
agama bentieensis Monard is shown to be a complex comprising two distinct species, one of 
which is described as new, the other sympatric with but unrelated to Agama agama (Linnaeus). 
Agama boensis Monard is not a valid form; the types are a mixed series of Agama sankaranica 
Chabanaud and Agama weidholzi Wettstein. 



Until a few years ago it was believed that only three species of Agama occurred in 
Nigeria: the ubiquitous Agama agama (Linnaeus) and the two smaller and less 
common species A . sankaranica Chabanaud and A . gracilimembris Chabanaud. But 
in 1963 Dr. Hilary Fry of the Ahmadu Bello University, Zaria drew my attention 
to the existence of a form which he described as being of similar adult size to Agama 
agama and as occurring on the smaller inselbergs in the Zaria region of northern Nigeria. 
The adult male was said to be distinguished by four indigo or black, clearly defined 
blotches on the throat. It was immediately recognized from an examination of an 
adult male which Fry submitted that this Agama was not conspecific with any of the 
three known species. With the co-operation of members of the staff of his depart- 
ment intensive efforts were made to obtain further examples as well as ecological 
data. The material resulting from these field studies contained more adult male 
blue throats '' but not adult females belonging to the same form, instead examples 
of Agama sankaranica, which in the field had been confused with the apparently 
undescribed form, and gravid females of yet another apparently undescribed agama. 
The absence in collections of females and immature individuals of this '' blue throat " 
continued until the following year when a Danish zoologist, Dr. Arne Schi0tz, who 
was working in the British Museum recognized the male blue-throated agama as 
being similar to ones he had collected in the Western Region. His large series, 

on inselbergs near Idanre and Igbetti, included both sexes and juveniles 
and were in the collections of the Zoologiske Museum Copenhagen. A detailed study 
there of his material enabled me to appreciate the difficulty that other collectors 
had encountered in the field in distinguishing the females and immature individuals 
from the superficially similar A. sankaranica and to elucidate the intraspecific 
variation of all the Nigerian forms of Agama. The study was extended by referring 
to type and other material in several institutions and to the British Museum's collec- 
tions including those acquired during the 1962 Northern Nigeria expedition. Finally 
the specimens on which Monard based the name Agama agama henueensis were 

ZOOL, 17, 3, 




The identity and relationships of Monard's Agama agama henueensts had not been 
questioned until Wermuth (1967) in a footnote pointed out that benueensis is probably 
no more than a local variety of Agama agama agama differing from it only in 
minute details of toe length and pattern. The original description of benueensis, 
extremely brief and inadequate, turned out to be based on a mixed series, some of 
the individuals being conspecific with the Nigerian '' blue-throated '' form, while 
others, strikingly different in their pattern and in the position of the nostril and with 
a lower average midbody count are closely related to Agama agama and apparently 

conspecific with a Nigerian population. The name benueensis is here applied to the 
'' blue-throat " form and a lectotype is designated. The information now available 
on the ecology, morphology, distribution and variation of Agama agama, A . benue- 
ensis, A . sankaranica and A . gracilimembris is given and a new species is described. 
Descriptions are based on the observed variation in the examined material. 
Throughout this paper the following abbreviations are used to refer to specimens 
or collections. 

B,M. British Museum (Natural History) 

M.H.N.C.F. Museum d'Histoire Naturelle, La Chaux de Fonds 
M.N. H.N. Museum National d'Histoire Naturelle, Paris 

Z.M.C. Universitets Zoologiske Museum Copenhagen 


The characters used in this study are defined as follows: 

Body scales. The shape, size and direction of imbrication of the trunk and flank 
scales, the extent of the keel relative to the axial length of each scale, from its anterior 
border to its apex. 

Microscopic structure of body scales. The number and position of the '' 

bearing " scale organs on the scales of the mid trunk region were studied by removing 
a series from the vertebral to the mid ventral row and mounting them according 
to the method described by Underwood (1957). 

Midbody scale count. The number of longitudinal scale rows around the middle 
of the trunk. 

Vertebral count. The number of transverse scale rows along the vertebral line 
from forelimbs (on a level with the axillae) to hind limbs (groin). 

Direction of imbrication of temporal and supratemporal scales. Direction of overlap 
of the scales in these regions, the regions being defined as : temporal, the area between 
the anterior rim of the ear and the postocular scales, overlying the temporal fossa 
and the squamosal and the posterior half of the jugal bones; supratemporals, the 
scales overlying the supratemporal arcade and the post orbital bone. 

Body length. The distance from the tip of the snout to the anterior margin of the 
vent, measured with dial calipers. 

Head length. The distance measured with dial calipers from the tip of the snout 
to the angle of the lower jaw at the posterior extremity of the articular. 

Upper labials. The number of scales bordering the upper lip, counted from the 
scale adjacent to the rostral to the corner of the mouth. 



Lower labials. The number of scales bordering the lower lip, counted from the 
scale adjacent to the mental to the corner of the mouth. 

Lamellae under fourth toe. The number of transversely enlarged scales under the 
fourth toe from the base of the claw to the junction of the fourth toe with the third 

Relative lengths of the toes. Comparisons of the first and fifth and of the third and 
fourth are expressed as the number of lamellae contained in the difference in lengths 
when the toes are laid flat and parallel to each other. 

Fig. I. Supratemporal and temporal areas of Agama; 

broken lines. 

g bones 


E — Eye 

J— Jugal 

M — Maxilla 

P — Parietal 

Po — Postorbital bone 

Q — Quadrate 

S — Squamosal 

St — Supratemporal arcade 

T — Temporal fossa 

Ty — ^Tympanum 


Cohn (1914) was the first person to report on the occurrence of dermal sense organs 
in Agama and he described in detail their histology as well as their external appearance 
and distribution on various parts of the body. His study was restricted to Agama 
colonorum {= A. agama). Subsequently Schmidt (1920) and Preiss (1922) examined 
the scale organs of other species and genera of lizard, Schmidt extending his studies 
on reptile receptors to the Gekkonidae, Iguanidae and Agamidae but limiting his 
investigation of Agama to an Asian species ^. sanguinolenta, which however exhibited 
several differences from A . agama, as described by Cohn, in the number and position 
of the organs on each trunk scale. It was not until 1937 and 1941 when Scortecci 
made an extensive survey of the lenticular and '' hair-bearing '' sense organs in the 
Agamidae that the density of these receptors and their positions on each scale were 
shown to have some value as a species character. (The agamid material which 




Scortecci used belonged to nineteen genera and eighteen species, his nine species of 
Agama all being African forms but none occurring in West Africa.) But despite the 
breakthrough afforded by Scortecci' s meticulous comparative studies the potential 
of the receptor character for helping to distinguish agamid species was passed by, 
yet Underwood (1957) in surveying the scale organs in pygopod lizards had stressed 
that they would repay further study and Wayne King (1962) had effectively demon- 
strated their usefulness in sorting out the taxonomy of Caribbean sphaerodactyl 
geckos. The only attention recently given to these scale organs is by Miller and 
Kasahara (1967) who have studied their innervation and who remark on the tactile 
function of the '' hair-bearing " receptor. 

In species whose mode of life is known correlations might be established between 
habits and the frequency with which the '' hair '' bearing organs occur, their position 
on the scales, their distribution on the body surface and the extent to which the 

bristles '' project beyond the edge of the scale. An investigation along these 
lines seems merited but unfortunately through lack of adequate field observations 
this could not be carried out on West African species of Agama. 

It has been found in the present study that the species number and the position 
of the " hair-bearing '' organs on the trunk scale do not vary ontogeneticaly, nor 
sexually, but that there is some individual variation which is discussed under the 
species. Nor does this character vary geographically, at least not in four of the 
Nigerian species where material from throughout their known ranges has been 
examined. Whether there is geographic variation in Agama agama receptors has 
still to be established but none is evident in the Nigerian specimens. 

apparently reliable taxonomic character not previously employed in 
keying out agamas is introduced into this study, namely the direction of imbrication 
of the temporal and supratemporal scales. These two characters, receptors and 
direction of imbrication, both of which are probably related to the animal's ecology, 
seem to have greater significance in classification than some of the characters so 
rigidly adhered to in the past in keys and descriptions of species of agama ; certainly 
the results of studying them in the Nigerian species are encouraging and an extension 
of study to other groups of agamid lizards may help elucidate the status and relation- 
ships of the many forms that have been described. 

Agama agama (Linnaeus) 
(Map, PL I, figs, c, d; Pis. 4, 5) 

Lacerta agama 'Linna.evis, 1758 : 207. 

Material examined. Nigeria, Sokoto Prov., i mile north of Kware, 15 miles 
north of Sokoto B.M. 1962. 1609; Katsina Prov., Rimi village, 13 miles South-east 
of Katsina B.M. 1962. 1608; Kano Prov., Birnin Kudu B.M. 1962.16x0; Bornu Prov., 
Maiduguri B.M. 1962.1611, Galtimare village, South of Maiduguri B.M. 1962.1612; 
Zaria Prov., Kawo, 4 miles north of Kaduna B.M. 1962.1596-98, Kujama, 25 miles 
South-east of Kaduna B.M. 1962.1599-1603, Zonkwa B.M. 1962. 1604 and 
Birnin Gwarri B.M. 1967.2207, Galma Fadama B.M. 1967.2208, Old City Zaria 


B.M. 1967.2209-2211, River Amethyst, Kaduna Road, Zaria B.M. 1967,2212; 
Plateau Prov., Jos B.M. 1962. 1613-54, 10 miles south of Jos on Miango Road B.M. 
1962.1593-95, Shendam B.M, 1962.1574-92; Oyo Prov., Igbetti Z.M.C. 
36567, 365691 36573. 36565, 36644, 36655, 36745. 36286-9; Ondo Prov,, 

Z.M.C. R36577-8, 36584, 36586. 
Description. Body subcyHndrical, little depressed above, between 3-3 and 3-8 

times the head length. The body of adult males rather triangular in section, the 
apex mid dorsal. Nostril on a line connecting the canthal supraciliaries with the 
rostral /first labial suture, that is on the canthal ridge; situated in the posterior half 
or third of a strongly swollen, pear-shaped nasal shield and directed upwards, or 
upwards and outwards. A keel usually present between the nostril and the anterior 
edge of the nasal. On its posterior margin the nasal forms a suture, generally a 
broad suture, with the first canthal supraciliary. The supraorbital scales and the 
scales overlying the frontal and anterior portion of the parietal quite smooth, the 
other scales on the head keeled, particularly so in juvenile specimens. The direction 
of imbrication of the scales overlying the supratemporal vacuity is forwards but the 
direction of the scales covering the postorbital bone alters to downwards and slightly 
backwards. The scales overlying the temporal fossa, imbricate in the reverse 
direction and between the two areas and along an oblique line, which marks the 
position of the squamosal and the posterior limb of the jugal, is a single row of quite 
strongly keeled juxtaposed scales (PI. I, fig. c and Text-fig i). Occipital enlarged, of 
variable size, its greatest width generally half to three quarters the diameter of the 
tympanum. Seven to twelve upper labials; a similar number of lower labials. A long, 
low nuchal crest of approximately fourteen scales but no dorsal crest. A poorly 
developed caudal crest present in adult males. The shape of the ear roughly a right- 
angled triangle, the tympanic membrane rather superficial. No conspicuous fringe of 
long, pointed scales along the anterior border of the tympanum but a few erect 
scales which in juveniles may be only slightly raised. A row of erect or somewhat 
spinous scales from the commissure of the lips to immediately behind and below the 
tympanum where it joins with one or two (in adults invariably two) round groups of 
spinose scales on the swollen area at the angle of the jaws. At the upper posterior 
corner of the tympanum a less elevated and more elongated group of erect scales is 
present. Midway between a point half way along the nuchal crest and the lower 
posterior corner of the tympanum lies a small oval group of spinose scales ; in front 
of and also behind this group a very small group of erect scales usually insignificant 
but if prominent the longest scales less than half the width of the tympanum. All 
these groups of erect scales are considerably more pronounced and the scales longer 
and more spinous in adult specimens, particularly in adult males, and in the juveniles 
the groups may be quite inconspicuous. A strong gular fold present in both sexes. 

Body scales homogeneous, those in the mid dorsal area flattened dorsoventrally, 
their distal margins rounded, often only feebly keeled and with short mucrones. 
Scales on the flanks more rhomboidal, particularly in adult males, the keels more 
pronounced and terminating in upward turning short mucrones ; distal margins some- 
what denticulate. Scales on either side of the nuchal crest small, barely mucrohate, 
approximately equal in size to those between the neck pit and ear. 



The sense organs of each trunk scale are situated at the distal margin or just below 
the dorsodistal edge of the scale. They are not confined to the area around the base 
of the mucrone but may extend along the entire distal aspect of the scale. No scale 
organ on a trunk scale has more nor less than one " bristle ". The number of 
receptors per scale varies from one to eight and while there is a tendency for the 
number to increase on the flanks and then diminish towards the venter no definite 
correlation could be established between the number of organs per scale and the 
longitudinal scale row although the body scale rows bearing the highest density are 
invariably between rows ten and sixteen. It is in this region of the trunk that the 
scales often have a rhomboidal shape with a somewhat denticulated distal margin 
and each organ is usually situated in an identation. Harris* (1963) statement that 
sense organs are absent from the ventral body scales is incorrect. Each ventral 
scale has one but no more than one '' hair-bearing " sensory organ at the apex. The 
smoother scales of the middorsum generally have one to four sensory organs per 

Cohn (1914) found only three organs on each of the scales of the back and describes 
the central one as placed somewhat to one side of the projecting tip of the keel. 
He reports a similar number of receptors on flank scales and says that the scales on 
the vertebral line may lack one or both lateral organs. He gives no indication of 
the origin of his material which he identifies as Agama colonorum, and there seems 
to be a strong possibility that a misidentification accounts for the much lower density 
of trunk scale organs in his specimens. 

Midbody scale counts of sixty-four Nigerian examples 59-77, the mean for thirty- 
five females being 68-2 and for thirty males 66-2; vertebral count 38-55. Mature 
males with eight to twelve, usually ten preanal pores. Eighteen lamellae under the 
fourth toe. First toe shorter than the fifth by two to seven (usually three to six) 
lamellae. Third toe shorter than fourth toe by one to two lamellae, very exceptionally 
subequal. Tail approximately twice the body length, depressed at the base and 
particularly so in adult males, its scales strongly keeled and mucronate and arranged 
in whorls. 

Intraspecific variation. Loveridge (1936) in discussing intraspecific variation 
in Agama agama remarked on the extraordinary wide range of midbody count, 
60-80, quoted by Boulenger (1885). Presumably Boulenger's counts were based 
solely on the material that was available in the British Museum at that time, namely 
twenty-one West African examples many of which bore no more precise locality 

" West Africa '\ and eight Angola and South West Africa examples, in other 
words a small sample from only a very Umited portion of the range of the species; 
In the 1936 publication Loveridge demonstrated without regard to any sex difference 
a gradual reduction in the same count from east to west (Uganda to Senegal) and 
in the belief that populations occurring west of Ghana have a considerably lower 
average count he suggested recognizing these individuals as belonging to a distinct 
race, savatieri Rochebrune. The recognition of a western race based solely on this 
character was ill conceived and hasty, particularly in view of Loveridge having at 
his disposal only one individual from Ghana and none from the 1,000 mile stretch 
between Ghana and the Cameroun. In 1956 Grandison in reporting on over one 



hundred Agama agama obtained in West Africa west of Ghana remarked on a south- 
west to northwest reduction from 68 to 60 in their average midbody count (range 
of the entire series 56-76). She assigned the individuals to A. agama savatieri, 
having overlooked the statement in Loveridge's later paper (1941) that Hallowell's 
name africana has priority over savatieri. 

(1961) counts (M 6i-6 (^ : 65-1 $) on twenty adult Liberian individuals 
which he identified as A. agama africana agree favourably with those given by 
Loveridge (1936, 1941) for Liberian material but suggest that Loveridge's samples 
in which the average number of scales round the body was sixty-two probably 
consisted mainly of males. 

The range of variation in midbody count in the sixty-five Nigerian examples 
studied here runs almost the whole gamut of the variation said to occur in africana 
but the average number of scales in males and females are intermediate between 
those quoted by Loveridge for Cameroun and Uganda individuals of the nominate 
form and those given by Loveridge, Daniel and Grandison for examples identified 
as africana and obtained west of Ghana. The Nigerian, Ghanaian and eighteen 
of the Senegal counts shown in the map are for adult specimens in the British Museum 

54-70(r» 25, m 60 

56-66fn 13, m 61 


60-60 (n 7<^^m 64) 
60-Sfrin9^ m 62-5) 

63-70fn 4, m 66) 

62-74(n 13, m^67) 


62-80(n 130*, m 70-4) 
68-74fn 79. *!?» 7V4j 

59-77(n 22<r, 
59-77 ( n 27$ 

74fn 7C 'm 7V4 
C_. I ^.' . 60-72(n 

57- 68(n 10 o*, m 61-6) 
58-72(n 10?, 

(n 35, m 66j 

62-74(n So*, m 70) 
68-74(n8$, m 70) 


6T^(n 13, m 6S) 



67-77 \t\ 13, m 

'4-84(r> 13, m'^79) 

Fig. 2 Geographic varation in midbody count in Agama agama. 

Thys van den Audenaerde (1963) in his study of Congolese species of Agama 
examined large samples of A , agama from the Congo and compared his findings with 
those of other authors for specimens from all regions in the range of the species. 
His thorough review of the pertinent literature available on Agama agama emphasizes 
the clinal sequence in midbody count from east to west although his argument is 
to some extent weakened by blind acceptance of identifications by others, to unequal 
sampling and to his not distinguishing between the counts for males and for females. 
He includes in his range of variation in A. agama a specimen from the Bauchi Plateau 
with ninety- two midbody scales that Grandison (1956) identified as A, sankaranica. 
While the individual is not available for re-examination it seems more than likely 
that it was incorrectly identified and is actually an example of A, benueensis, a 
species closely related to sankaranica and in its juvenile livery resembling sank- 
aranica but with a considerably higher midbody count. An Air specimen with a 



low count of 58 was also mentioned by Grandison (1956) but excluded from the table 
of means; it too was probably incorrectly identified and may be an example of the 
large scaled species Agama boueti Chabanaud. This species was described from Gao 
on the Niger. Examination of type material shows that its trunk receptors are 
similar in number and position to those of A. hihronii Dumeril of North West 
Africa and are quite unlike any other West African form. 

Thys van den Audenaerde maintains that development of the keel and mucrone is 

body count, the lower the count the stronger the keeling, but I can 
establish no such Unk in the Nigerian material where variation in this character 
seems to be related to the proximity of sloughing. He further comments on the 
geographical variation that occurs in A . agama colouration, in particular of the head, 
throat and tail, which is perhaps correlated with the cHnal gradient in scaling and 
he draws attention to Angolan intermediates between the nominate race and the 
Angolan race mucosoensis. A similar cline in the development of the nuchal crest 
which both Stejneger (1893) and Thys van den Audenaerde (1963) have suggested 
exists, has not yet been effectively demonstrated. Much more quantitative data 
on large samples from throughout the range of the species would have to be compared, 
sex with sex, if a true picture is to emerge. Meanwhile the only East African figures 
available are those of Stejneger based on only six Kenya examples, and of Barbour 
and Loveridge (1928) for eighteen adult Usambara specimens and in neither case 
is the sex or the proportion of the sexes stated. Since in Nigeria at least, there is a 
sex difference in the length and height of the crest, the spines being fewer and shorter 
in the females, the East African figures are not necessarily comparable with those 
for twenty adult male Nigerian individuals (13-17 : M 14-9). 

Thus the abundance of races of A. agama that have been described reflects not 
the occurrence of legitimate geographically localized subdivisions but accidents of 
sampling or more likely the work of splitters. There would appear to be no evidence 
of geographic isolates but only of marked individual variation within a population 
and Thys van den Audenaerde's conclusion that Agama agama is a large polymorphic 
species rather than a polytypic species would seem to be the logical one. 

Colour. Immature individuals and adult females are similar in colouration. 
In their reproductive colour phase the head is a medium brown or brown with green 
tinges, the back fawn or brown with darker brown marks of variable shape and 
position but often in the form of rather indistinct diamond-shaped outlines along 
the vertebral region. The upper part of the head, the temporal region and the neck 
have pale green spots and short longitudinal stripes of the same colour. The 
undersurfaces are a grey-white. Mature males that are dominant over territories 
are characterized in their reproductive colour phase by an orange head and neck, 
indigo blue body and by a median orange segment on a black-tipped tail. Other 
colour phases occur and these as well as the reproductive colour phase are admirably 
and fully described by Harris (1964). 

Ecology. In Nigeria wherever the savanna and forest have been disturbed by 
human agency, bridges built, houses and shacks erected, trees felled, rubbish dumped, 
Agama agama seems to have established territories. It is very much a commensal 


of man and is common around native houses, farm buildings, litter, wood piles and 
along paths. The species is also common in the boulder strewn country flanking 
the road from Jos to Miango. Schi0tz (personal communication) says 
habitat preferences do not include flat uniform savanna, nor large rocks, nor dense 
forest but as soon as such areas acquire human habitations or are in other ways 
disturbed by man Agama agama takes up residence. While this is largely true, the 
rocky mass of Kujama Hill south of Kaduna supports a number of Agama agama yet 
on the bare rock surfaces of the vast gneiss domes 1,000-3,000 ft. high at Igbetti 
and Idanre Agama agama is absent, its place being taken by Agama benueensis. 

Range. Senegal to Ethiopia, southwards to Tanzania and Angola. 

Agama paragama sp. n. 

(PL I, figs, a, b; Pis. 4, 5) 

Agama agama bemieensis Monard 1951 : 131 (part). 

Material examined. Holotype: Nigeria, Zaria Prov., \ mile Southwest of 
Ahmadu Bello University, Zaria City, collected by C. H. Fry from Parkia tree on 
17 . ii . 1964, B.M. 1967 . 2215 $. 

Paratypes: Nigeria, Zaria Prov., same locality as holotype B.M. 1967. 2216, 
7, Zaria Road B.M. 1967. 2214, Mile 5, Zaria Road B.M. 1967. 2213, Ahmadu 
Bello University campus B.M. 1967. 2217, no information, probably in vicinity of 
Zaria City B.M. 1967. 2218, Birnin Gwarri, 62 miles Southwest of Funtua B.M. 1968. 

497, Zaria City B.M. 1968.482-483, Veterinary Unit, Zonkwa B.M. 1962. 1605, 

B.M. 1962. 1607; BornuProv., Potiskum B.M. 1968.498, Yo B.M. 1968.496; Plateau 
Prov., Naraguta, Jos B.M. 1968.484-485. Cameroun, Bangouve M.H.N.C.F. 

1162, Rei Bouba M.H.N.C.F. 1432, 1440, 1459, Mayo Sala M.H.N.C.F. 1374, 

Ngaouyanga M.H.N.C.F. 965 (six of the paralectotypes of Agama agama benueensis.) 

These individuals differ from A . agama not only in body size, proportions, greater 
spinosity and in lower average midbody and vertebral counts but in an absence of 
sexual dimorphism in size. They resemble A . agama in the position of the nostril being 
on the canthal ridge and in having multiple sense organs on the distal edges of the 
scales of the ilanks and the two forms are undoubtedly closely related but being 
sympatric in Nigeria specific status is given to paragama. 

Relatively few examples of paragama have been collected and the form is so far 
known only from northern Cameroun and from the Northern Region of Nigeria, 
although one specimen obtained by C. H. Fry in a crevice 10 ft. up a Terminalia 
tree in the Botanical Gardens, University College, Ibadan was thought by him to 
belong to this form; the specimen has not been seen by me. Although situated in 
the high forest belt Ibadan has many clearings that have the character of cultivated 
savanna woodland and support typical savanna species. Schi0tz (1963) describes 
the vegetation of the University College campus as '' dry forest and culture savanna ". 
It is possible that where suitable habitats are available this species, like Agama 
benueensis, penetrates the Western Region's forest belt. 



Description of Holotype. Body subcylindrical, depressed above, 3-6 times the 
head length. Nostril on the canthal ridge, situated in the posterior half of a swollen, 

and directed laterosuperiorly. Nasal keeled from its 
anterior tip to the upper rim of the nostril, forming a suture on its posterior margin 
with the first canthal supraciliary. The supraorbital scales and the scales overlying 
the frontal and the anterior portion of the parietal quite smooth, remaining head 
scales keeled. The direction of imbrication of the supratemporal scales is forwards 
and slightly upwards; anteroventrally the scales overlying the postorbital bone 
face downwards and backwards. The scales immediately in front of the ear, that 
is those overlying the temporal fossa, imbricate in the reverse direction to those 
covering the lower postorbital and between the two areas and along an oblique line 
marking the position of the squamosal and posterior Umb of the jugal is a single row 
of large, keeled, juxtaposed scales. Occipital enlarged, its greatest width equal 
to three-quarters the horizontal diameter of the tympanum. Upper labials 8 ; g 
(one divided) ; lower labials 9 : 10. A short but conspicuous nuchal crest formed 
from nine, long, conical scales, the longest equal to three-quarters the width of the 
tympanum; no dorsal nor caudal crest. Ear triangular. On anterior margin of 
ear a fringe of two to three long, pointed scales, on upper posterior border one long, 
pointed scale surrounded by several erect scales (this group barely developed on 
left side of head). At the posterior extremity of the lower jaw, that is behind and 
below the ear, a small, round group of spinose scales of which one scale is three to 
four times longer than the others. A similar group immediately above. Between 
a point midway along the nuchal crest and the lower, posterior corner of the tymp- 
anum lies a prominent round tuft of long spinose scales the longest scale three- 
quarters the tympanic diameter; in front (right side only), and also behind, a small, 
round but insignificant group of erect scales. (The lengths of the scales comprising 
the nuchal crest and the groups of spinose neck and ear scales are considerably longer 
and more pronounced in male specimens: see note on variation.) A well developed 

Body scales homogeneous, the longitudinal rows converging towards the midline, 
scales on the dorsum and flanks equally strongly keeled and mucronate ; the mucrones 
of the flank scales rather broad at their base, tapering rapidly and rising at an angle 
to the scale. Distal margins of the body scales rounded on each side of the mucrones. 
Ventral scales smooth. Scales on either side of the nuchal crest with 
compressed keels and long, almost erect, mucrones, the scales at least twice the size of 
those between the neck pit and ear. 

The maximum number of sense organs on the middorsal and flank scales is nine, 
the minimum three. They are situated just under the dorsodistal margin and may 
extend as far as the '' shoulders " of the scale. No scale organ on a trunk scale has more 
nor less than one " bristle '\ The highest density of receptors on midtrunk scales 
occurs on longitudinal rows 3-5. Each ventral scale bears a receptor at its apex. 

Midbody scale count 60, vertebral count 30, eighteen lamellae under the fourth toe. 
First toe four or five lamellae shorter than the fifth toe, third toe half to one lamella 
shorter than the fourth. 

Tail one and a quarter times body length (tip of tail missing), depressed at base, 


only very slightly compressed distally, its scales keeled and strongly mucronate 
and arranged in whorls. 
Length of body 94-0 mm. 

Variation. The following variation in meristic characters in the paratype series 
was noted. Nigerian individuals: midbody scale count 50-66 (M58-5), vertebral 
count 26-34 (M 30-1), number of lamellae under the fourth toe seventeen to twenty- 
one, first toe shorter than fifth toe by between three and five lamellae, third toe half to 
two, usually one to two lamellae shorter than fourth toe. Differences in the develop- 
ment of the rosettes of neck and ear spines and in the lengths of the scales forming 
the nuchal crest are attributable to sex and maturity, in adult males the longest of 
the pointed nuchal crest scales being as much as equal to the width of the tympanum, 
likewise the most prominent scales in the neck tufts, whereas in females they are 
less than three-quarters the diameter of the tympanum. The number of scales form- 
ing the nuchal crest varies from eight to ten, the usual number being eight. The 
density and distribution of '' hair-bearing '' receptors on the midbody scales of the 
paratypes is similar to the holotype. Sex dichromatism is discussed below. The 
tail of the adult male is considerably more compressed than that of the female and 
has a slight crest ; the base of the tail is depressed but not to the same extent as in 
Agama agama. 

Juvenile A. paragama in which the neck and ear rosettes and nuchal crest are 
insignificant can be readily distinguished from young individuals of A. agama by 
the relative size of the scales above and below the central neck rosette and by the 
roundness of the rosette. Further distinguishing features are the pattern, lower 
vertebral count, gradual tapering of the tail and less depressed base to the tail. 

All the Cameroun examples, except an adult male (Mayo Sala M.N.H.C.F. 1374), 
fall within the range of variation in midbody count of the Nigerian individuals but 
they have a shghtly higher number of transverse rows on the trunk, their vertebral 
count being 36-38. The Mayo Sala example with seventy- two midbody scale rows 
is well outside the normal range of variation in the species yet in its other morpho- 
logical characters it agrees with paragama. 

Colour. Male. In life the upper surfaces of the body and limbs of the adult ^ 
are a metallic blue with white flecks. On the back the flecks tend to be arranged 
in narrow transverse lines, no more than one scale in width, with approximately 
eight lines between fore and hind limbs. The vertebral zone is whitish, the head 
chalk white, the proximal two- thirds of the tail whitish, the tail tip black. Except 
for the whitish areas the general appearance of the ^ is rather similar to the reproduc- 
tive colour phase of a male Agama agama. Harris (1964) describes in detail the 
various colour phases of both <^ and $ Agama agama and while he refers to a geograph- 
ical variation in the colour of the head of adult males, northern individuals (Kano) 
having a deep yellow while more southern populations have either orange or vermilion, 
he makes no mention of any adult male agama in which the orange or yellow tones 
are replaced by white. 

Female. None of the colour phases described by Harris for female A, agama 
occurring in Nigeria corresponds to, or in any way approaches the colour pattern 



of female A. paragama which is characterized by four or five pairs of more or less 
clearly defined, large, round, brown spots on a lighter brown (orange in life) background 
on the dorsum between fore and hind limbs* In those specimens in which the spots 
are ill-defined the lateral halves of the spots are indicated by narrow, dark brown 
outhnes. The flanks and the upper surfaces of the limbs and tail are of a similar 
shade of brown to the spots; the head is mottled brown. In both sexes the throat, 
and in some specimens also the sides of the neck and chest, has a dark network on a 
cream ground which takes the form of isolated, round, cream spots and is dissimilar 
to the gular pattern of A. agama which tends rather to consist of a longitudinal 
arrangement of dark Hues or blotches (Harris, 1964) . The distinctness of the gular 
spotting does not appear to bear any relation to the maturity or sex of the individual, 
for example one gravid female has a pronounced pattern while in another a pattern 
is barely discernible. No pattern whatsoever is present on the throat of one adult 
male but a distinct network is present in another mature male 

Adult female paragama are larger and considerably more robust than female 
agama. Three gravid female paragama vary in body size from 96-107 mm. while 
the gravid female agama examined in this study vary from 86-91 mm. Harris 

records 97 mm. as the average body length of sixty-eight Nigerian adult 
female agama and Chapman & Chapman (1964) regard all Ghanaian female -4. agama 
of about 90 nam. as mature. Adult male paragama are considerably smaller 
108 mm.) than mature male agama obtained in the same general area (107-137 

Ecology. Little information on the ecology of paragama is available. Fry 
(in litt.) states that the species is to be found on the trunks and lower limbs of shade 
trees and collectors have suggested that it occupies a different ecological niche 
from A . agama but only the specimens collected at Mile 7 Zaria, Potiskum and Birnin 

i bear information 
; the Yo specimen, 
examined material 1 

(Mango and Parkia 

Range. West Africa; Northern Region of Nigeria to northern Cameroun. 

Agama benueensis Monard 
(PL 2, figs, a-d; Pis. 4, 5) 

Agama agama benueensis Monard, 1951 : 131 (part). 

Agama agama boensis Monard, 1951 : 130 (part: males). 

} Agama sankaranica: Grandison, 1956 : 231 (part: Bauchi specimen). 

Material examined. Cameroun, Upper Benue, Ngaouyanga M.H.N.C.F. 
956-7, Sakdje M.H.N.C.F. 1269, Mayo Sala M.H.N.C.F. 1373, Rei Bouba M.H.N.C.F. 
1442, 1460, 1517 (seven of the paralectotypes of ^. a. benueensis), Ngaouyanga 
M.H.N.C.F. 961 (designated lectotype of A. a. benueensis). Nigeria, Oyo Prot;., 

Igbetti Z.M.C. R 36342, 36345, 36347-51. 36353, 36389. 36392, 36394, 36397, 36399, 
36401, 36599, 36601-12, 36614-20, 36661-2, 36702-3; Ondo Prov., Orosuta, Idanre 

Z.M.C. no number, Idanre Z.M.C. R 36576, 36579, 36580-2, 36585, 36588-9; Zaria 


Prov., R. Amethyst, Kaduna Road B.M. 1967.2234-8, B.M. 1967.2263, Ahmadu 
Bello University campus, Zaria B.M. 1967.2239-44, Siberia, A.B. University campus 
Zaria B.M. 1967.2254-7, Old City, Zaria B.M. 1967.2245-7, B.M. 1967.2258-61, 
Mile 5, Zaria Road B.M. 1967.2248-52, B.M. 1967.2253, Zaria B.M. 1967.2262, 
Near Oomawa, at foot of Kwatarakwashi rock B.M. 1968.490-495; Plateau Prov., 
Jos B.M. 1968.486-487, Hillcrest School, Jos B.M. 1968.488. 

Remarks. The whereabouts of four of the type specimens of Agama a. benueensts 
(959, 1372, 1441 and 955) listed by Monard (1951) are not known at present. Thys 
van den Audenaerde tells me he believes that part of the Monard material originally 
lodged in the herpetological collection at La Chaux de Fonds was transferred to other 
European institutions and that the missing Cameroun agamas may have been among 
it. The thirteen individuals from the benueensts type series that are available and 
have been re-examined belong to two morphologically very different forms, and as 
referred to in the introduction to this paper, the name bemieensis is applied to the 
blue-throated form. The other form is described as A. paragama sp. n. The 
Monard specimens here assigned to A. benueensts are those males from the Upper 
Benue that are described by him (195 1) as having '' une tache noiratre a la gorge '', 
'' une tache foncee a la gorge '' and '' une raie vertebrale claire '', as well as four 
females and a juvenile male that have a feebly developed pattern characteristic of 
immature and female Nigerian benueensts, and two males from Tibati, Adamaoua 
Plateau which he assigned to A. agama boensis. The Tibati males, M.H.N.C.F. 673, 
784, have been examined by Dr. Braes trup who reports (in litt.) that both individuals 
have chin as well as gular spots. The presence of the gular spots, even if faint, 
identifies them as benueensts for this species is the only West African agama that 
has a bilobed blotch in this region in the adult male. In the Tibati males the midbody 
count is about 74, the vertebral counts 45 and 47 which accord with Cameroun 
examples of benueensts. Neither I nor Dr. Braestrup has seen the female, M.H.N.C.F. 
726, which Monard also referred to A. a. boensis and its identification remains in 
doubt . 

Diagnosis. A large sized Agama (body length of c^ up to 113-0 mm., of $ to 
74-0 mm.) related to A. sankaranica and characterized by a high average number of 
midbody scale rows, 74-98 (M 88 ; N 71) in Nigerian populations, ventrally directed 
temporal scales and marked sexual dimorphism. Dorsal trunk scales each with 
one to three single '' hair-bearing '' scale organs close to base of mucrone, nostril below 
can thai ridge; in these respects similar to A, sankaranica. 

Description of Lectotype M.H.N.C.F. 961, ^. Body subcylindrical, little 
depressed above, 3-7 times the head length. Nostril below a line connecting the 
canthal-supraciliaries with the rostral-labial suture, its aperture round, situated in 
the posterior third of an elongated, pear-shaped nasal, directed superior-laterally. 
From the upper rim of the nostril to the front of the nasal shield is a distinct ridge. 
Canthal-supraciliaries 8:9, the anterior forming broad sutures with the nasals at 
their posterior borders. Supraorbitals quite smooth, supratemporals distinctly 
keeled, temporals only very slightly keeled. The scales overlying the supratemporal 
arcade and the postorbital bone directed forwards and slightly upwards; ventrally 




the imbrication of the scales covering the posterior part of the jugal, the broad 
portion behind and below the eye, as well as those over the temporal fossa directed 
downwards and slightly backwards; no row of keeled juxtaposed scales, 
plate enlarged, broadly oval, its length equivalent to the horizontal diameter of the 
tympanum. Seven to ten upper labials, eight to twelve lower labials. Tufts of 
enlarged, conical, rather flat scales around the ear, one, consisting of three scales, 
on the anterior border, one on the upper posterior border, on the lower posterior 
border a group of about four to five more prominent conical scales and below this 
group a curved row of about five flattish conical scales. At the side of the neck 
midway between the tympanum and a shallow neck pit, but a little above, a rosette 
of spines, the longest scale in the rosette two-thirds the length of the nasal. ^ Behind 
and a little below this rosette and situated above the neck pit an oblique or vertical 
line of conical scales. Above and slightly behind the tympanum a prominent 
erect scale surrounded by a few lower but erect scales. Ear subtriangular, its 
horizontal diameter equal to the distance from the tip of the snout to the middle 
of the nasal. A gular pouch present. Nuchal crest low, formed from about ten 
conical scales; no dorsal nor caudal crest. Dorsal scales imbricate, homogeneous, 
strongly converging on the vertebral line, strongly keeled but the keel not markedly 
compressed, mucrone finely tapering, ventrals and subcaudals smooth. 

A maximum of two scale organs on each body scale but usually only one and 
situated under the dorsodistal edge of the scale close to the base of the mucrone. 
Each scale organ bears a single '' hair 

Midbody scale count 78; vertebral count 46; preanal pores in two transverse 
rows, 13 anteriorly, 15 posteriorly. The adpressed hind limb reaches to posterior 
border of the tympanum. Nineteen lamellae under the fourth toe. First toe three 
or four lamellae shorter than fifth toe, third toe one lamella shorter than fourth toe. 

Tail slightly depressed at the base; broken. Caudal scales keeled and mucronate, 
arranged in distinct whorls. 

Body size 94-0 mm. 

Colour in alcohol. Apart from a conspicuous whitish vertebral line approxi- 
mately three scales wide and extending from the nape to the base of the tail there is 
little colour pattern discernible. Any pattern that may have been present on the 
flanks of the living animal has become obscured by preservation and sloughing and 
only a slight mottHng of cream and dark brown is now evident. The throat bears 
a large, blackish, bilobed patch at the level of the gular fold; in front of this patch 
and extending on to the chin are several small, dark spots on a cream background. 
The rest of the undersurfaces are a grey-cream. 

Colour in life of Nigerian examples. Dorsum grey or brown with whitish 
dark-edged ocelli on the flanks which are arranged in two or three longitudinal rows, 
the spots in the outer, or middle in the case of individuals with three rows on each 
side, generally larger than the others. Across the back from the shoulders to the 
tail a series of dark chocolate brown bands which, particularly in juveniles, are 
considerably broader and often diamond-shaped on the vertebral line and enclose a 

narrow, longitudinal, whitish streak. The streaks become confluent in some 


males and females (see sex dimorphism) and form a diffuse yellowish vertebral stripe, 
Schi0tz (personal communication) describes the entire dorsum and belly of adult 
males as having a distinct bluish tinge, but not of such an intense blue as the Agama 
agama dominant male, while the chin and throat are yellow with two to four dark 
blue or black patches. The upper surface of the male's head is irregularly mottled 
green; the tail is greenish with indistinct yellow bands. 

The ground colour of the female varies according to the colour of the rocks or 
soil in its habitat. At maturity the transverse bands on the trunk are reddish brown 
and the sides of the head are greenish. The ventral surface of both the female and 
the j u venile is white but the chin and gular region of most individuals has a network 
of dark lines forming irreg 

Intraspecific variation. An analysis of meristic variation in the material 
examined shows some geographical differences. The samples from the western 
region (Igbetti and Idanre) have the highest midbody and vertebral counts, midbody 
84-98 (M 91-3 : N 34), those from the Jos Plateau and the Cameroun the lowest, 
68-82 (M 76 : Nil) while the gradation in the Zaria Province population almost 
spans these two extremes, 74-96 (M 85-8 : N 34). The variation in these characters 
for the entire series is : midbody 68-98 (M 87-0 : N 79), vertebral 41-56 (M486 : 
N76). The correctness of the identification of the individuals at the extremes of 
the variation is not in doubt, a combination of other characters such as throat mark- 
ings, position of nostril, trunk scale organs and direction of imbrication of the post- 
orbital/temporal scales assigning them unquestionably to henueensis. Variation in 
the number of fourth toe lamellae (nineteen to twenty-three) and in the relative 
lengths of the first and fifth, and third and fourth toes is independent of the variation 
observed in midbody and vertebral counts and appears to have no geographical nor 
ontogenetic significance. The third toe may equal the fourth but generally it is 
shorter by half to one lamella. Variation in the relative lengths of the first and fifth 
toes is greater; the first is always shorter than the fifth but the difference may be as 
many as four lamellae. 

In the majority of the examples studied the lateral head scales between the eye 
and the ear all imbricate downwards or downwards and slightly backwards but in a 
few individuals, juveniles as well as adults, the scales in the lower posterior quarter 
of this area, that is those overlying the temporal fossa are directed forwards and 
upwards, in other words in a similar direction to those of Agama agama] however 
they are not as strongly keeled as in agama nor are they separated from the post- 
orbital/supratemporal scales by an oblique row of juxtaposed scales. 

Scales on either side of the nuchal crest are not or only sHghtly mucronate and are 
only slightly keeled. They are considerably larger than the scales below the central 
group of spines on the side of the neck. 

Sexual dimorphism and dichromatism. Some sexual dichromatism is evident, 
notably in the conspicuous gular and chin spots which are a characteristic feature 
of the adult male. The twin gular spots develop before the chin spots and first 
appear as a darkening of the median gular region which later becomes an intense 
black or navy bilobed blotch; occasionally the lobes are separated on the midline. 

ZOOL. 17, 3. 





The gular spots do not develop until the male reaches a body size of over 73 mm., 
although there is one exception among the fifty-six males examined, a 697 mm. 
individual collected at Igbetti. By 80 mm, most of the males have developed two 
additional dark spots; these are on the chin and somewhat oval and parallel to each 
other on either side of the median Une, The development of these gular and chin 
spots in adult male henueensis does not seem to be seasonal; they are present in 
Nigerian individuals collected in January, February, March, June, July and August. 
The three largest males in the Cameroun series (72-5, 78-3 and 74-0 mm.) which were 
obtained in July and in September have only gular spots. 

Adult males retain some evidence of the juvenile pattern; generally the two longi- 
tudinal rows of ocelli along each flank are discernible, although feebly, as well as 
remnants of the five dark transverse lines in the mid dorsal region on each side of the 
midline from shoulders to groin but the light coloured vertebral streaks, a character- 
istic feature of the juvenile livery, invariably coalesce and form an ill defined light 
vertebral zone extending from the nape to the tail. The female sample is insufficient 
to be certain whether a similar light vertebral zone always develops at maturity. 
Only six of the fifteen females are gravid and of these six only the two Cameroun 
examples (M.N.H.C.F. 1460, 1269) have a clearly defined continuous light line. 
In the other examples the vertebral pattern is obscure in all but one of the Igbetti 
examples (Z.M.C. 36351). In both sexes the dorsolateral series of ocelli becomes 
progressively less distinct as the individuals mature. Egg bearing females were 
obtained in July, August and September. Preanal pores are usually ten to twelve 
in a single series but in three Idanre males, the only male henueensis collected there, 
and in the lectotype the pores are in double series 13 + 14, 10 + 13, 8 + 13, 13 + 15 

Ecology. In the western region of Nigeria Schi0tz (personal communication) 
says that A. henueensis is strictly confined to large rocky outcrops such as the vast 
gneiss inselbergs where it is common both in the savanna (Igbetti) and in the forest 
(Idanre). Other ecological niches that he searched in southern Nigeria, such as 
mounds of boulders, small rocky areas, produced no examples. He noted that 
henueensis hides in rock crevices and among vegetation at the perimeter of an insel- 
berg but conducts its feeding and sexual activities on the actual surface of the rock. 
Farther north on the Jos Plateau the species is much less abundant but it is again 
quite common in Zaria Province on the inselbergs around Zaria City as well as in a 
variety of other habitats such as on mounds of laterite on open farmland in the 
Galma Fadama, a locality some miles from the nearest inselbergs, on sandy river 
banks (Amethyst River) and in sandy stream beds (R. B. Walker, personal communi- 
cation) . Near Sabon Gari , in an area of scrap cars and garbage tips , both A . henueensis 
and A. agama were taken and no specific habitat preference was noted but Schi0tz 
maintains that at Igbetti A. henueensis, A. sankaranica and A. agama can be seen 
virtually only a few feet from each other yet each still adhering to its preferred 


A careful search by Schi0tz of the Shai Hills in Ghana and suitable localities in the 
Eastern Region of Nigeria failed to produce any henueensis and he believes the species 


may also be absent from the Bamenda region of the Cameroun. Agama benueensis 
seems to be endemic to the Western and Northern regions of Nigeria and the upper 
Benue valley in the Cameroun. 

Monard (195 1) does not record the biotopes for the type series of benueensis and 
he describes only the general areas around the villages. The i ; 1,000,000 maps of the 
Upper Benue show the villages where benueensis were taken as being in an area 
deeply dissected by the Benue and its tributaries and at altitudes varying from 
740 ft. to 1,400 ft.; only Ngaouyanga is on a high steep-sided wooded platform, 
the other localities are in or at the sides of valleys in woodland savannah. 

Range. West Africa: Western and Northern Regions of Nigeria and northern 


Agama sankaranica Chabanaud 
(PL 3, figs, a, b; Pis. 4, 5) 

Agama sankaranica Chabanaud 191 8 : 105, 

Agama boensis Monard 1940 : 155 (part: adults only). 

Material examined. Portuguese guinea, Madina Boe M.H.N.C.F. 865, 867-8 
(three of the syntypes of Agama boensis). Republic of guinea, Beyla M.N. H.N, 
21.309-316, Kankan M.N. H.N. 21.297-298, Kerouane M.N. H.N. 21.299-308, 
B.M. 1921.11.12.1-5. Mali, Moussaia, Sankaran M.N. H.N. 1901.395 (holotype of 
Agama sankaranica). Ghana, B.M. 1930 .6.9.8 Korley Bu, Accra B.M. 1931 .5.6.3, 
B.M. 1932. 6. 1. 5-7, Tamale B.M. 1927.9.27. 199, Tafo B.M. 1962.910. Nigeria, 
Benue Prov., Makurdi B.M. ig^y .12.4.14; Plateau Prov., Shendam B.M. 1962. 1573; 
Kano Prov., Kano B.M. 196T.2069, B.M. 1962.567 (no. 567 later skeletonized); 
Zaria Prov., B.M. 1967.2206, Rimi, South-east of Kaduna B.M. 1962. 1572, Zonkwa 
B.M. 1962. 1571, Ahmadu Bello University campus, Zaria B.M. 1967.2205, 14 miles 
South of Zaria B.M. 1967.2204; Lagos Prov., Lagos B.M. i960. 1.7. 28; Oyo Prov., 
Igbetti Z.M.C. R 36386, 36388, 36390, 36393, 36395-6, 36398, 36403, 36405, 36407. 
36597. 36607, 36697-8, 36691, 36758. 

Diagnosis. A medium sized Agama (body length of adult males averaging 
66 mm., of mature females 76 mm.) belonging to the group of agamas that have the 
nostril below the canthal ridge, homogeneous scaling and dermal sense organs not 
exceeding three in number on each trunk scale. Related to Agama benueensis 
Monard but distinguished by a lower average midbody count (69-6) and by dorsally 
directed temporal scales. 

Description. Body subcylindrical, scarcely depressed above, 3-3*7 times the 
head length. Nostril below the canthal edge, situated in the posterior half of a 
slightly swollen, oval nasal, directed laterally. No post nasal separating the first 
canthal supraciliary from the nasal. Scales of the head keeled; the interorbital 
scales considerably smaller than the supraorbitals ; imbrication of temporals directed 
upwards and of supratemporals directed upwards and forwards (PI. 3 fig. b) 
Occipital large, its greatest width equivalent to or as much as one and a half times 
greater than the diameter of the tympanum. Eight to twelve upper labials; a 

200L. 17, 3. 




similar number of lower labials. Usually three tufts of long, conical spines on pos- 
terior border of ear; behind and above these a single group of spines and yet farther 
back and just above a shallow neck pit another group of somewhat shorter spines, 
A weakly developed gular pouch, A low nuchal but no dorsal or caudal crest. 
Dorsal scales homogeneous, broadly rounded, strongly keeled and mucronate, the 
sharp high keel extending the length of each scale, the mucrone very slender. Scales 
on the flanks slightly smaller than dorsals, similarly keeled and mucronate. Ventrals 
keeled only in juveniles. *' Hair-bearing '' sensory pits on the dorsal midtrunk 
scales confined to around the base of the mucrones, at or just under the dorsodistal edges 
of the scales. The maximum number of scale organs on each scale is three but 
invariably no more than two are present. In the vertebral region at midbody there 
are generally two '' hair-bearing '' pits, each with a single hair-like projection, the 
one pit close to the base of the mucrone or under the mucrone and concealed in dorsal 
view, the other on the opposite side of the mucrone but near the apex of the scale. 
On more lateral scales there is a tendency for only the more lateral, that is lateral 
with respect to the mucrone, to develop. 

The number and position of the scale organs on the dorsal scales bear a close 
similarity to the condition obtaining in East African rueppelli and described and 
figured by Scortecci (1937). Midbody scale count 64-78 (M 69-6 : N 61); vertebral 
count 32-46 (M 38-8 : N 61). Mature males with eight to twelve usually ten preanal 
pores in a single transverse row. Fifteen to twenty-one lamellae under the fourth 
toe. First and fifth toes subequal or first toe one lamella shorter. Third toe usually 
subequal to fourth toe, rarely half to one lamella shorter or longer than fourth. 
Tail one and three quarter times the body length, depressed at the base, slender and 
cylindrical distally, its scales strongly keeled and very distinctly mucronate, not 
arranged in distinct whorls. 

Colour in alcohol. Upper parts reddish brown. Invariably a conspicuous 
yellowish or russet vertebral stripe from behind the head to the proximal third of the 
tail, much broader and always prominent in the cervical region but narrowing rapidly 
to a width of approximately two scales on the rest of the body. On each side of this 
stripe and from nape to tail broad, dark brown, transverse bands, six between the 
nape and hind limbs. At the lateral extremities of these bands light-coloured ocelli 
or elongated blotches which in some specimens coalesce to form a light dorsolateral 
stripe. In some examples the vertebral stripe widens at the level of each transverse 
band to form light diamond-shaped areas. Halfgrown individuals occasionally 
have indistinct marbling on the flanks below the ocelli or dorsolateral stripe. In 
one mature male, collected at Zaria, the transverse bands are broken up into irregu- 
larly shaped blotches. Between the eyes a thin dark transverse line is present; 
four similar lines radiating from the eyes to the lips are usually evident. 

indistinctly cross-barred with narrow dark lines. 

There are no obvious external differences between the 
females and immature males. Adult males have eight to twelve preanal pores and 
also broken, longitudinal, blue stripes on the chin and throat which become more 
prominent with maturity and often enclose a large, dark blue, median gular spot 


and a lozenge-shaped chin blotch. The rosettes of neck spines and the lips may also 
be blue in mature males. Adult females are on average rather larger than adult 
males. The body length of mature females varies from 68-85 i^^- (average 76-2 mm.) 
while males vary from 62-76 mm. (average 66*4 mm.). 

Remarks. The examined material consists of examples from throughout the 
known range of the species. No evidence of geographical variation can be demon- 
strated. The examples from the extreme west of the geographical range of sank- 
aranica are the three adults, two males and one female, in the series of syntypes of 
Agama boensis Monard. These individuals unquestionably belong to sankaranica. 
The throats of the males have the blue blotches and stripes that are typical of mature 
male sankaranica, furthermore the scale counts and other morphological characters 
including the density and distribution of trunk receptors agree well with sankaranica 
although one male with a midbody count of 64 is exceptional in having 20 ; 21 
lamellae under the fourth toe, which is outside the range for all other specimens 
examined (fifteen to nineteen). The fourth syntype of boensis, a juvenile, lacks a 
nuchal crest and clusters of neck spines and has a single apical scale organ on each 
body scale; it is conspecific with Agama weidholzi Wettstein. (Grandison, in 
press.) The Cameroun individuals assigned by Monard (1951) to ^. a. boensis are 
discussed in the section on A. benueensis. 

Ecology. In the field this species has often been confused with female and imma- 
ture Agama benueensis and as a result some ecological information claimed to be on 
A. sankaranica has had to be disregarded or treated with suspicion in this study. 
Dr. Arne Schi0tz' reliable field notes and comments make it clear that he readily 
distinguishes sankaranica from benueensis and his notes correspond with my own 
field observations in stating that unlike all other Nigerian species of Agama, with the 
exception of gracilimembris , sankaranica does not form family groups but is a solitary 
creature. The species has been observed and captured in ploughed farmland and 
in maize and cassava plantations when running along the rutted ground, also on 
paths in grass covered savanna. Schi0tz' notes state that he saw the species on the 
ground only in grass covered savanna or in rather dense tree savanna and never in 
forest nor on rocks. The localities at which examples of A. sankaranica have been 
collected are, with one exception, in the Doka and True Guinea woodland belts from 
Nigeria westwards to Portuguese Guinea; the exception is an individual taken at 
Kano which is towards the southern limit of Sudan woodland. 

Range. West Africa: Portuguese Guinea to Nigeria. 

Agama gracilimembris Chabanaud 

(PI. 3, figs, c, d; PI. 6) 

Agama gracilimembris Chshansiud, 191 8 : 106. 

Material examined. Dahomey, M.N. H.N. 04.114-5 (syntypes). Nigeria, 


Shendam B.M 
Samaru Bush. 





African Rep., ''Pays des Senoussi '' M.N. H.N. 17. 191 (according to information 
received from Dr. J. Guibe this specimen was collected in the neighbourhood of 
NdeUe, 8^ 25' N : 20^36'E). 

Diagnosis. A small sized Agama (body length of males up to 47 mm., of females 
to 57 mm.) lacking tufts of erect spiny scales behind the ear and bearing only one 
receptor on each body scale, related to Agama weidholzi Wettstein but distinguished 
by heterogeneous body scales, strongly keeled head scales and a low fourth toe 
lamellar count (13-14). 

Description. Body sHghtly depressed, 3-3-6 times the head length. Nostril 
below the canthal edge, situated in the posterior half of a convex, often keeled, oval 
or pear-shaped nasal; directed laterally. Interorbital scales as large or larger than 
the supraorbitals. Above the nasal a continuous series of three clearly defined, 
somewhat elongated scales runs from the rostral to the canthal-supraciliaries. The 
first canthal-supraciliary is separated from the nasal by a small postnasal. Scales of 
the head strongly keeled, somewhat rugose ; imbrication of temporal scales directed 
downwards and slightly backwards. Occipital large, its greatest width equivalent 

to the diameter of the tympanum. Nine to twelve, usually ten upper labials; 
eight to eleven, usually ten lower labials. No tufts of spines around the ear, instead 
single, short, conical scales close to the border of the ear and three to four tubercles 
on the side of the neck in a slightly curved row extending from the upper edge of 
the ear to above a point midway between the neck pit and the arm insertion ; addi- 
tional tubercles irregularly scattered below this row and between the ear and the 
neck pit, the lowest at the jaw angle. No gular fold. No trace of a nuchal or dorsal 

. Dorsal scales heterogeneous, those in the vertebral region a little larger 
than the lateral series; on the flanks, and irregularly disposed, some large scales 
which are as big and occasionally bigger than the vertebrals. 

The scales of the midtrunk region broadly oval or pear-shaped with pronounced 
'' shoulders ", dorso ventrally flattened on each side of a high keel, the edges of which 
rise abruptly from the scale. The keel extends half to two-thirds the length of the 
scale but does not project beyond the scale apex to form a mucrone. A single sensory 
organ is present on the dorsodistal surface of each of the midbody scales just under the 
keel tip or very slightly to one side and sometimes evident only when the scale is 
viewed from behind. Generally each of these organs bears one but never more 

than one '' hair '', which projects slightly beyond the apex of the scale and by macro- 
scopic inspection may be mistaken for a mucrone. Gular and ventral scales strongly 

Midbody scale count 70-85 (M 74-8 : N 16) vertebral count 30-46 (M 37 : 
N 15). Mature males with eight to twelve preanal pores in a single transverse row. 
The adpressed hind limb reaches to the tympanum. Thirteen to fourteen lamellae 
under the fourth toe. The first toe is usually either one lamella longer than the 
fifth toe or equal to it; rarely (two examples) is it shorter (half lamella). The third 
toe exceeds the fourth toe in length by half to two lamella. Tail one and a half 
times the body length, depressed at the base, slender and cylindrical distally its 
scales strongly keeled and slightly mucronate, and not arranged in whorls. 
Body length of gravid females 49-0-56-7 mm., of mature males 44-o-47'0 mm. 



Comparison of characters of Nigerian species of Agama 


sankaranica gracilimembris 

Dorsal body 

Position of 

.. nostril 

Dorsal body 
scales keeled 
and mucronate 


Homogeneous Homogeneous Homogeneous Homogeneous Heterogeneous 

On canthus 



(M68-2 $ : 

M66-2 ^) 

Vertebral count 38-55 

Nuchal crest 

Imbrication of 
scales in 
temporal region 
on left side of 

Number of 
receptors per 

dorsal scale 

Keeling of 
dorsal head 

Fourth toe 

First to fifth toe 
expressed in 
number of 

Third to fourth 







or < (1-3) 

Body length 
. of gravid $ 

Body length 
of adult cj 

Tufts of spines 
behind ear and 
on side of 

84-91 mm. 

107-137 mm. 


On canthus Below canthus Below canthus Below canthus 

+ (strongly) 









96-107 mm. 

99-108 mm. 



M85 • 8 Zaria) 

26-34 (M30) 42-56 (M50) 



(rarely \ 

1-2 (rarely 





65-74 mm. 


80-113 'Tim 


+ (strongly) Keeled but not 



(M69 • 6) 

32-46 (M39) 



1-2 (rarely 



or<(i) or 
68-85 mni. 

62-76 mm. 




(M74 • 8) 

46 (M37) 









49-57 mm. 

44-47 mm 

* Based on Nigerian individuals only. 



Colour in alcohol. (^ Upper surfaces of body and limbs greyish brown; usually 
a clearly defined light vertebral stripe which extends from behind the occiput to the 
base of the tail. Straddling the vertebral stripe from nape to base of tail are nine 
<> shaped, dark chocolate brown marks which in some specimens are rather 
obscure. lu the same region and alternating with these marks, diamond-shaped 

brick orange zones occasionally present. The 

snout, supraorbital and 

tympanic and parietal regions invariably dark brown or blue black but a lighter 
area over the occipital is always present. Dark, irregular, longitudinal lines are 
usually present on the throat, chest and belly but they may extend no farther back- 
wards than the gular region. In a sexually mature male these lines are dark brown 
and closely set to give a brownish appearance to the entire belly; the throat has a 
dark median area. 

5 The entire dorsal surfaces of the body, tail and limbs brown with obscure, irreg- 
ularly placed lighter areas. An ill defined light vertebral stripe occasionally present. 
Snout, supra-orbital area, cheeks and olbique stripe from below eye to commissure 
of lips dark brown; occipital, temporal and nuchal regions light fawn. Above and 
behind each ear a large ill defined blackish patch in which the small pointed scales 
stand out clearly as light bluish or greyish spots. Lower surfaces greyish white with 
faint longitudinal dark lines from throat to belly. In some specimens these lines 
do not extend farther than the chest. 

Ecology. A. gracilimemhris is the rarest of the West African savannah species 
and has been recorded from only a few localities all of which are in the Doka and 
True Guinea woodland vegetation belts and on the extreme southern limit of Sudan 
woodland (at Kano). Its biotope has not been described. 

Range. West Africa : Dahomey to Ubangi Chari. 


The stimulating enthusiasm, interest and collecting ability of R. B. Walker and 
Dr. C. H. Fry of the Ahmadu Bello University is gratefully acknowledged and thanks 
are extended to Prof. A. P. Mead for his cooperation. Thanks are also due to Dr. 
G. T. Dunger who co-operated in obtaining samples from the Jos area. I am 
especially indebted to Dr. A. Schi0tz, Danmarks Akvarium for graciously making 
available his field notes on his extensive collections and for valuable discussions and 
hospitality during my Copenhagen visit. Dr. F. W. Braestrup, Copenhagen Museum 
has not only taken a keen interest in this study but has been a source of encourage- 
ment and assistance. To Professor J. Guibe, Paris Museum, Dr. V. Aellen, Geneva 

and Dr. R. Matthey, La Chaux de Fonds I am also indebted for the loan of speci- 
mens, for supplying information on specimens of Agama in their care or for museum 
facilities. I wish to thank too A. F. Stimson, Paula D. Jenkins and Belinda A. Brindley 
for their help in scale counting and in preparing slides, skeletons and line drawings. 

Summary of taxonomic changes : 

Agama boensis Monard part 

A . sankaranica Chabanaud 

Agama boensis Monard part (juvenile) = A. weidholziV^eiisitin 
Agama a. benueensis Monard part 

Agama a. benueensis Monard part 

A . benueensis Monard 
A . paragama sp. n. 



Barbour, T. & Loveridge, A. 1928. A comparative study of the herpetological faunae of 

the Uluguru and Usambara Mountains, Tanganyika Territory, with descriptions of new 

species. Mem. Mus, Comp. Zool. Harv. 50 (2) : 87-265, 4 pis. 
Chabanaud, p. 191 7. Enumeration des reptiles non encore Studies de TAfrique occidentale, 

appartenant aux Collections du Mus6um, avec la description des especes nouvelles. BidL 
. Hist. nat. Paris, 23 : 83-105, 13 text figs. 
1918. fitude compl6mentaire de deux Agama de TAfrique occidentale et description de 

quatre especes nouvelles de reptiles de la meme region. BidL Mus. Hist. nat. Paris, 24 : 

Chapman, B. M. & Chapman, R. F. 1964. Observations on the biology of the lizard Agama 

agama in Ghana. /. Zool. Lond. 143 : 121-132, 6 figs. 
CoHN, L. 1914. Die Hautsinnesorgane von Agama colonorum. Zool. Anz. 44 : 145-155, 

7 text-figs. 
Daniel, P. M. 1961. Notes on the life history of Agama agama africana (Hallowell) in Liberia. 

Spec. Pubis. Ohio herpet. Soc. no. 3 : 1-5, 2 tables. 
Grandison, a. G. C. 1956. On a collection of lizards from W6st Africa. Bull. Inst. fr. Afr. 

noire (A) 18 : 224-245, 4 text-figs, 2 maps. 
1969. Agama weidholzi (Sauria: Agamidae) of West Africa and its relationship to Agama 

gracilimembris . Ibid, (in press). 
Harris, V. A. 1963. The anatomy of the Rainbow Lizard. Hutchinson Tropical Monographs, 

London: 1-104, 39 text-figs. 
1964. The life of the Rainbow Lizard. Hutchinson Tropical Monographs, London: 

1-174, 39 text-figs, 
Keay, R. W. J. 1959. Vegetation map of Africa south of the Tropic of Cancer. Oxford 

University Press, London. 
King, W. 1962. Systematics of lesser Antillean lizards of the genus Sphaerodactylus. Bull. 

Fla. St. Mus. biol. Sci. 7 : 1-52, 17 text-figs. 
Linnaeus, C. 1758. Systema Naturae. Tenth Edition: 824 pp. 
Loveridge, A. 1936. African reptiles and amphibians in Field Museum of Natural History. 

Pubis. Field Mus. nat. Hist. Zool. ser. 22 : i-iii. 
1 94 1. Report on the Smithsonian-Firestone Expedition's collection of reptiles and 

amphibians from Liberia. Proc. U.S. natn. Mus. 91 : 1 13-140, i text-fig. 
1952. Mission A. Villiers au Togo et au Dahomey (1950) XII. Tortoises and lizards. 

Bull. Inst. fr. Afr. noire (A) 14 : 229-242. 
Miller, M. R. & Kasahara, M. 1967. Studies on the cutaneous innervation of lizards, 

Proc. Cal. Acad. Sci. 34, 16 : 549-568, 20 text-figs. 
MoNARD, A. 1940. R6sultats de la Mission scientifique du Dr. Monard en Guin^e Portugaise 

(i937"i938)- VIII Reptiles. Archos. Mus. Bocage, 11 : 147-180, 2 pis., 5 text-figs. 
1951. R6sultats de la mission zoologique Suisse au Cameroun. Mem. Afr. noire, 

I : 1-244, 13 figs. 

Preiss, F. 1922. Uber Sinnesorgane in der Haut einiger Agamiden. Jena Z. Naturw. 
58 : 25-76, 9 text-figs. 

Rochebrune, A-T, de, 1884. Faune de la S6n6gambie. Reptiles. Octave Doin, Paris, 
89 : 1-22 1, 20 pis. 

ScHi0TZ, A. 1963. The Amphibians of Nigeria. Vidensk. Medd. dansk. naturh. Foren. 
125 : 1-92, 28 text -figs., 4 pis. 

Schmidt, K. P. 1919. Contributions to the herpetology of the Belgian Congo based on the 
collection of the American Congo Expedition, 1 909-191 5, BtdL Am. Mus. nat. Hist. 
39 : 385-624, 27 figs., 32 pis. 

Schmidt, W. J. J920. Einiges iiber die Hautsinnesorgane der Agamiden insbesondere von 
Calotes, nebst Bemerkungen iiber diese Organe bie Geckoniden und Iguaniden. Anat. Anz. 
53 : II 3-1 39, 16 text-figs. 


ScORTECCi, G. 1937. ^li organi di senso della pelle degli Agamidi. Memorie Soc. ituL Set. 
nat, 10 (11) : 159-206, 39 text-figs., 2 pis. 


I recettori degli agamidi. Memorie Soc, ital. Sci. nat. 10 (m) : 209-326, 80 text- 

Stejneger, L. 1893. On some collections of reptiles and batrachians from East Africa and 
the adjacent islands, recently received from Dr. W. L. Abbott and Mr. William Astor 
Chanler, with descriptions of new species, Proc. U.S. natn. Mus. 16 : 711-741. 

Thys Van Den Audenaerde, D. F. E. 1963. Les Agamidae du Congo: les especes et leur 
distribution g6ographique. Rev. ZooL Bot. afr. 68 : 203-256, 4 text-figs. 

Underwood, G. 1957. On lizards of the family Pygopodidae. A contribution to the morph- 
ology and phylogeny of the Squamata. /. Morph. 100 (2) : 207-268, 11 text-figs. 

Wermuth, H. 1967. Liste der rezenten Amphibien und Reptilien: Agamidae. Tierreich, 
86 : 1-127. 

Wettstein, O. 1932. Eine neue Eidechse aus Senegambien. Zool. Anz. 99 : 303-305, i 



a. Agama paragama sp. n. Holotype B.M. 1967 . 2215. 

b. Agama paragama sp. n. Lateral view of the head of the holotype. 

c. Agama agama L. Lateral view of the head of a juvenile female B.M. 1967 ,2212. 

d. Agama agama L. Dorsal view of same individual. 

Bull. Br. Mus. nat. Hist. (Zool.) 17, 3 


ZOOL. 17, 3 



a. Agama benueensis Monard. Dorsal view of an immature male B.M. 1967.2239. 

b. Agama benueensis Monard. Dorsal view of an adult male B.M. 1967.2234, 

c. Agama benueensis Monard. Throat of a sexually mature male B.M. 1967.2253. 

d. Agama benueensis Monard. Laterial view of the head of an immature male B.M. 1967 . 2239. 
Note developing dark gular patch . 

Bull, Br. Mus. nat. Hist. (Zool.) 17, 3 


ZOOL. 17, 3; 



a. Agama sanharanica Chabanaud, Dorsal view of an adult female B.M. 1967.2204. 

b. Agama sankaranica Chabanaud. Lateral view of the head of the same individual. 

c. Agama gracilimembris Chabanaud. Lateral view of the head of an adult male B.M. 1 96 1 . 949 

d. Agama gracilimembris Chabanaud. Dorsal view of same individual. 

. Br. Mus. nat. Hist. (Zool.) 17, 3 



Stereoscan electron microscope photographs of dorsal midbody scales. Arrows point to the 

sense organs. 

Upper row left to right. Agama agama 

Agama par agama sp. n. 

Lower row left to right. Agama benueensis 

Agama sankaranica 

B.M. 1962.1595 X 48 
Holotype X 28 

B.M. 1967.49 X 52 
B.M. 1967.2204 X 47 

. Br. Mus. nat. Hist. (Zool.) 17, 3 



Stereoscan electron microscope photographs of the apices of the same trunk scales as illustrated 
in Plate 4 showing the positions of the sense organs relative to the keel and dorsodistal margin 
and the single ** hair '* protruding from each receptor. 

Upper row left to right. Agama agama X 70 

Agama paragama X 70 

Lower row left to right. Agama henueensis X 131 

Agama sankaranica X 260 

Bull. Br. Mus. nat. Hist. (Zool.) 17, 3 



Stereoscan electron microscope photographs of a dorsal midbody scale of Agama gracilis 
membris B.M. 1961 .949 showing the position of the single ** hair-bearing " sense organ directly 
below the termination of the keel. 

X no. Right X 580, 

Bull. Br. Mus. nat. Hist. (Zool.) 17, 3